101 geodynamic modelling: how to design, interpret, and communicate numerical studies of the solid Earth

van Zelst, Iris; Crameri, Fabio; Pusok, Adina E.; Glerum, Anne; Dannberg, Juliane; Thieulot, Cedric

doi:10.5194/se-13-583-2022

Articles | Volume 13, issue 3

https://doi.org/10.5194/se-13-583-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Collection:

Celebrating 15 years of Solid Earth interactive open-access...

https://doi.org/10.5194/se-13-583-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 13, issue 3

Review article

| Highlight paper

|

17 Mar 2022

Review article | Highlight paper |

| 17 Mar 2022

101 geodynamic modelling: how to design, interpret, and communicate numerical studies of the solid Earth

Iris van Zelst, Fabio Crameri, Adina E. Pusok, Anne Glerum, Juliane Dannberg, and Cedric Thieulot

Download

Final revised paper (published on 17 Mar 2022)
Supplement to the final revised paper
Preprint (discussion started on 18 Feb 2021)
Supplement to the preprint

Interactive discussion

Status: closed

CC1:
'Comment on se-2021-14', Paul PUKITE, 20 Feb 2021

Just a comment that will help distinguish this document from what would be generic information that could be applied to any scientific discipline. What geophysics, astrophysics and similar disciplines have in common is the inability to set up a controlled experiment in the lab, which is an important factor in many other disciplines. Nowhere in the document is the lack of experimental control mentioned, which has an impact on how model validation is done. So can't test out a model hypothesis by setting up an experiment in the lab (e.g. no scaling of gravity possible, therefore any tidal forcing models are impossible to do in a lab). What this means is that the concept of cross-validation becomes much more important, in contrast to the typical hypothesis testing and prediction that are the usual yardsticks for evaluating the utility of a model. In addition, can't wait for predictions on geological time-scales that will unfold over the course of years to millenia, but do have historical data that is amenable to cross-validation analysis.
See https://en.wikipedia.org/wiki/Cross-validation_(statistics)
BTW, I come from a solid-state physics background where controlled experiments rule over everything else, and so this gaping hole in the discipline related to geophysics is glaringly obvious. This should be taken as advice from the trenches, no more than that.

Citation: https://doi.org/10.5194/se-2021-14-CC1
- AC1: 'Reply on CC1', Iris van Zelst, 15 Sep 2021
  
  Thanks for this comment. This is indeed a good point and we incorporated it into the manuscript by adding a line to the introduction (l 53-54) and adding a line at the end of Section 6.3 - Testing the model against observations:
  “Considering this lack of experimental control from both the real world and analogue models from the lab (see Section 4), techniques such as probability analysis and cross-validation of the models beyond the typical testing and predicting of hypotheses become increasingly important to assess how well models capture the real world.”
  
  Citation: https://doi.org/10.5194/se-2021-14-AC1
RC1:
'Comment on se-2021-14', Paul Tackley, 12 May 2021

This is an "educational" paper; it is not presenting new science or techniques, but instead is introducing newcomers to geodynamical modelling. As such, it is very good, and I have only some corrections and suggestions detailed in the attached report.

Citation: https://doi.org/10.5194/se-2021-14-RC1
- AC2: 'Reply on RC1', Iris van Zelst, 15 Sep 2021
  
  Thanks for the extensive and detailed comments. They have greatly improved the manuscript! You can find our detailed response in the supplement.
  
  Citation: https://doi.org/10.5194/se-2021-14-AC2
RC2:
'Comment on se-2021-14', Boris Kaus, 01 Jun 2021

This is a lengthy, but well written and useful manuscript that highlights many aspects of geodynamic modelling. I have a number of minor suggestions to further strengthen the manuscriot in the attached pdf.

Citation: https://doi.org/10.5194/se-2021-14-RC2
- AC3: 'Reply on RC2', Iris van Zelst, 15 Sep 2021
  
  Thanks for the extensive and detailed comments. They have greatly improved the manuscript! You can find our detailed response in the supplement.
  
  Citation: https://doi.org/10.5194/se-2021-14-AC3
RC3:
'Comment on se-2021-14', Laurent Montesi, 23 Jun 2021

This manuscript serves as a well-designed guide for modeling mantle and crustal-scale processes. As is necessarily the case with an exercise of the sort, it does in place reflect the personal experience and opinions of the authors, but it overall does a very nice job of remaining neutral, and even the more engaged sections are full of important information that will be useful for moving the field forward. I particularly appreciated the discussions of figure accessibility and Section 9 “Software, data, and resource management”. There are also some very important discussions of the objectives of modeling, in particular the difference between “specific” and “generic” modeling. Both are presented as equally valuable, which is an important message to pass to both modeling and non-modeling communities.

Although the paper is already well organized and well written, I do have a few suggestions that could lead to significant rewriting, as detailed in the attached PDF. These include 1) make clearer that the focus of the paper is on geodynamic modeling of the mantle and the crust. 2) emphasize the importance of hypotheses in the model design process. 5) emphasize the role of constitutive relations 6) explain how brittle failure can be implemented. 7) better contrast FEM and FDM. 10) clarify what an “unphysical behavior” is. My biggest concern, though is with section 8.1 “Structure of a geodynamic modelling manuscript” that I find overly prescriptive. Other comments can be seen as minor.

Please note that most of my suggestions can be seen as a matter of personal preference and should not stand in the way of the publication of the paper (even 8.1!). The authors have produced an important manuscript that will many gain a better understanding of geodynamic modeling. That said, there is room in our discipline for personal preferences, and I welcome continued discussion of the topic. More of us should take an occasional pause from the pace of research to reflect on the higher objectives of our work, in this case geodynamic modeling. This paper is a great way to get the conversation started. Thank you for putting it together.

Citation: https://doi.org/10.5194/se-2021-14-RC3
- AC4: 'Reply on RC3', Iris van Zelst, 15 Sep 2021
  
  Thanks for the extensive and detailed comments. They have greatly improved the manuscript! You can find our detailed response in the supplement.
  
  Citation: https://doi.org/10.5194/se-2021-14-AC4

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload

AR by Iris van Zelst on behalf of the Authors (15 Sep 2021) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (20 Sep 2021) by Taras Gerya

RR by Boris Kaus (06 Dec 2021)

RR by Paul Tackley (17 Dec 2021)

RR by Laurent Montesi (17 Jan 2022)

Suggestions for revision or reasons for rejection

Once again, it was a pleasure to read this important and well-organized manuscript. The authors have generally changed the manuscript where appropriate in response to my comments. In a few cases, they disagree with my suggestions and gave clear explanations why. As I wrote in my original review most comments express personal opinions or preferences, and so I am satisfied with the authors’ response. I do have a couple of further requests for clarification and wish to return to some of our discussions.

• The title has been updated to mention communication, which is a great addition and emphasizes a particularly original part of the paper. However, the authors declined my suggestion to mention the mantle, crust, or lithosphere in the paper, as that distinction is made in the abstract. While that is true, I still feel that readers from different geodynamic subdisciplines may be frustrated by the emphasis on convection and could use a clarification of the context of the paper when they first meet it, i.e., when they read the title. Elements of discussion (e.g., writing, study organization) are universal, but a large part of the paper discusses specific equations. Someone working on magma dynamics (two-phase flow), groundwater flow (Darcy’s law), or even earthquake mechanics (elastodynamics) would be frustrated with the appearance that geodynamics is assumed to be equivalent to mantle convection. You refer to Schubert et al. (2001) for the derivation of the equations, but please note that this book, although very important, is focused on mantle convection, not geodynamics in general. Your response to my concern of Eq. 15 is that it is the most common additional equation in geodynamics models. This ignores every problem based on elasticity or fluid transport. It is your paper, and you have the right to frame it as you wish. It just is not as general as you might think.
• In response to my comment to line 415, you say that the two criteria are equivalent for an “incompressible, 2D, plane strain model”. That does not make sense to me. The failure criteria only involve stress. Incompressibility and plane strain assumptions involve strains, so unless you assume a rheology, there is no general link between the conditions you express and the failure criteria. Even in 2D, Mohr-Coulomb features corners, as it is based on the maximum differential stress, while Drucker-Prager is smooth, as it is based on invariants (as you added to section 2.2.4).
• The following should not lead to a change in the paper, but I would like to push you a little on the topic of non-uniform color scales. Note that I am a fan of your color scale and try to use them in all my recent work, so, fundamentally, I agree with you. However, the effect of using a color scale that highlights a certain value is not very different from one of the “alternatives” you mention, superposing a contour. The key is that the person using a non-perceptually uniform scale should do it with a purpose (most use those scales without giving it a thought, and that’s bad). It is not easy to use these scales properly, so I agree that there is no need for going through the trouble, but I personally prefer to leave options open.

Minor comments on the updated text:
• Line 54: there have been approaches for gravity scaling using centrifuges (e.g., Ramberg, H., (1967) Model Experimentation of the Effect of Gravity on Tectonic Processes, Geophysical Journal International, 14, 307–329, doi:10.1111/j.1365-246X.1967.tb06247.x )
• Figure 3: I just noticed that the drawing for a plastic material is actually brittle. We use a plastic approximation for frictional sliding, where, unlike that snapping bar, the materials stay in contact after failure.
• Lines 208-210: I am confused by the discussion of a “viscous isentropic relaxation time scale”. If you discuss “viscous relaxation”, how can “elastic forces” dominate the time scale? A quick look at Curbelo et al. (2019) paper shows that the time scale does involve viscosity but no elastic moduli. Please include the reference to back up these statements for a viscoelastic body and the time scales shown on line 210. By the way, it may be worth quickly mentioning that fluid transport can lead to additional volume changes (viscous compaction; see McKenzie 1984, Ricard et al. 2001, Wilson et al., 2014a).
• Line 235: Without a rheology, the stress tensor does not contain information on the deformation of the material. That would make it a deformation vector.
• Line 381: Very confusing sentence. “Rocks can deform elastically, by brittle failure” implies that “brittle failure” enables elastic deformation, and of course, it’s not the case.

Hide

ED: Publish subject to minor revisions (review by editor) (17 Jan 2022) by Taras Gerya

AR by Iris van Zelst on behalf of the Authors (23 Jan 2022) Author's response Author's tracked changes Manuscript

ED: Publish as is (27 Jan 2022) by Taras Gerya

ED: Publish as is (27 Jan 2022) by Susanne Buiter (Executive editor)

AR by Iris van Zelst on behalf of the Authors (31 Jan 2022) Manuscript

Short summary

Geodynamic modelling provides a powerful tool to investigate processes in the Earth’s crust, mantle, and core that are not directly observable. In this review, we present a comprehensive yet concise overview of the modelling process with an emphasis on best practices. We also highlight synergies with related fields, such as seismology and geology. Hence, this review is the perfect starting point for anyone wishing to (re)gain a solid understanding of geodynamic modelling as a whole.